Predicting alpha-synuclein pathology by REM sleep behavior disorder diagnosis

Inability to accurately diagnose Lewy type alpha-synudeinopathy (LTS) pre-mortem has been a major obstacle to clinical care and research. Probable REM sleep behavior disorder (PRBD) diagnosed with support of instruments such as the Mayo Sleep Questionnaire (MSQ) may provide a cost effective means of predicting LTS. Since 2007, 602 subjects in the Arizona Study of Aging and Neurodegenerative Disorders had clinician assessment for PRBD (298 with, 304 without support of the MSQ), completed cognitive and movement examinations, and had neuropathological assessment. Mean age at death was 84.8 years. Histological evidence of LTS was found in 80/101(79.2%) cases with PRBD and 198/501 (39.5%) without PRBD (p < 0.001). Overall sensitivity for predicting LTS by PRBD diagnosis was 28.8%, specificity 93.5%, positive predictive value (PPV) 79.2%, negative predictive value (NPV) 60.5%. Diagnosis of PRBD was less frequently present in subjects without LTS [4/105 (3.8%) of healthy controls, 42/255 (16.5%) AD, 2/33 (6.1%) progressive supranuclear palsy (PSP) without LTS] than in subjects with LTS [11/46 (23.9%) DLB, 58/104 (55.8%) PD, and 4/16 (25.0%) PSP with LTS.] PRBD was not present in any of 46 subjects with incidental Lewy body disease (ILBD). MSQ-supported diagnosis of PRBD appears useful for predicting LTS in manifest neurodegenerative disease, but not necessarily ILBD. Additional prospective autopsy research, including well-characterized polysomnogram-confirmed RBD subjects, is needed to elucidate the earliest tissue abnormalities in the idiopathic (premotor/pre-dementia) stage of RBD.